Hartmann Scheiblechner

from Wikipedia, the free encyclopedia

Hartmann Scheiblechner (born August 16, 1939 in Graz ; † December 24, 2010 ) was an Austrian psychologist. From 1972 to 2004 he was Professor of Psychology in the Department of Psychology at the Philipps University of Marburg .

Life

Hartmann Scheiblechner studied psychology and anthropology at the University of Vienna from 1959 to 1964 . Then he was a scholarship holder and assistant at the Institute for Advanced Studies in Vienna in the Department of Sociology until 1968 . After receiving his doctorate in Vienna in 1968 with an experimental study on the formation of associations, he became a research assistant in the Department of Methodology and Mathematical Psychology at the University of Vienna, whose head, Gerhard H. Fischer, is a recognized expert on probabilistic models of measurement in psychology. In 1972 Scheiblechner was appointed professor for psychological methodology at the Department of Psychology at the Philipps University of Marburg. Hartmann Scheiblechner was still scientifically active after his retirement in 2004.

Scientific achievement

The starting point of his scientific work is the statistical model of the Danish statistician Georg Rasch for answering questions in performance tests ( Rasch model ). The model postulates: The probability that a person will give a correct answer to a test question is a function of two parameters: the difficulty of the task and the person's ability. The greater the skill of the person compared to the difficulty of the task, i. H. the greater the numerical difference between the person parameter and the task parameter, the greater the probability that the person will give the correct answer. Rasch assumes a logistical function for this connection. Scheiblechner's early work continues the logistic approach (linear logistic model, logistic model for latency times, logistic model for sociometric elections).

His main work, however, is the development of a non-parametric item response theory (non-parametric IRT). In this, he dispenses with the very restrictive assumptions of the logistic models and only assumes that people and items can be classified in terms of ability or difficulty (greater / less / equal relationships instead of numerical parameters). He postulates a doubly monotonous functional relationship between the probability of a solution, the person's ability and the difficulty of the task. He developed different variants of the model: ISOP, ADISPO and CADISOP for different types of items. Items with two answer categories (right / wrong) with more than two answer categories (e.g. ordered levels of correctness) and multi-dimensional models. He starts from an axiomatic approach from which he develops the model, but also deals with practical questions of application: How to check whether the data corresponds to the model and what the test evaluation should look like in order to get the ranking from the data to appreciate the people and the items optimally.

A Dutch group with the statistician RJ Mokken followed a similar approach as Scheiblechner.

Fonts

  • The separation of individual and system influences on behavior in social contexts . In: Acta Psychologica . tape 35 , 1971, p. 442-460 .
  • Personality and system influences on behavior in social contexts: Frequency models . In: Acta Psychologica . tape 36 , 1972, p. 322-336 .
  • Specifically objective stochastic latency mechanisms . In: Journal of Mathematical Psychology . tape 19 , no. 1 , 1979, p. 18-38 .
  • Isotonic ordinal probabilistic models (ISOP) . In: Psychometrika . tape 60 , no. 2 , 1995, p. 281-304 .
  • Corrections of theorems in Scheiblechner's treatment of ISOP models and comments on Junker's remarks . In: Psychometrika . tape 63 , 1998, pp. 87-91 .
  • Additive conjoint isotonic probabilistic models (ADISOP) . In: Psychometrika . tape 64 , 1999, pp. 295-316 .
  • Nonparametric IRT: Testing the bi-isotonicity of isotonic probabilistic models (ISOP) . In: Psychometrika . tape 68 , no. 1 , 2003, p. 79-96 .
  • A unified Nonparametric IRT Model for d-Dimensional Psychological Test Data (d-ISOP) . In: Psychometrika . tape 72 , no. 1 , 2007, p. 43-67 .
  • Rasch and pseudo-Rasch models: suitableness for practical test applications . In: Psychology Science Quaterly . tape 51 , 2009, p. 181-194 .
  • H. Scheiblechner, R. Lutz: The construction of an optimal one-dimensional test using non-parametric test theory (NIRT) using the example of the MR SOC . In: Diagnostica . tape 55 , 2009, p. 41-54 .

Web links

Individual evidence

  1. ^ Kubinger, Klaus D .: In Memoriam Hartmann H. Scheiblechner . In: Psychological Test and Assessment Modeling . tape 55 , no. 2 , April 1, 2013, ISSN  2190-0493 ( questia.com [accessed December 8, 2017]).
  2. ↑ Prabook. Retrieved December 8, 2017 (American English).
  3. ↑ Marburger UniJournal No. 35 • January 2011 • ISSN  1616-1807
personal data
SURNAME Scheiblechner, Hartmann
BRIEF DESCRIPTION Austrian psychologist
DATE OF BIRTH August 16, 1939
PLACE OF BIRTH Graz
DATE OF DEATH December 24, 2010